Jesús M. Hernández-Mangas

Improved binary collision approximation ion implant simulators

An efficient binary collision approximation (BCA) ion implant code with good prediction capabilities for semiconductor materials (Si, GaAs, SiC) with only one fitting parameter for low implantation doses is presented. It includes specific interatomic potentials and recent improvements in physical models for inelastic stopping. A periodic ab initio full bond electron density for the target is used. Damage accumulation is supported using a modified Kinchin–Pease model [G. H. Kinchin and R. S. Pease, Rep. Prog. Phys. 18, 1 (1955)]. Also, some of the BCA integration algorithms and target selection procedure have been refined. An algorithm commonly used for statistical noise reduction has been modified to also improve the noise reduction in the lateral and shallow zones. The agreement with experiments is good, even under channeling conditions and for different target materials. A comparison with experimental secondary ion mass spectroscopy results for several projectiles and targets is presented.

Atomistic Monte Carlo simulations of three-dimensional polycrystalline thin films

An atomistic Monte Carlo code to simulate the deposition and annealing of three-dimensional polycrystalline thin films is presented. Atoms impinge on the substrate with selected angular distributions, and grains are nucleated with different crystalline orientations, defined by the tilt and rotation angles. Grain boundaries appear naturally when the islands coalesce, and can migrate during both deposition and annealing simulations. In this work we present simulations of aluminum films. We examine the influence of the temperature, deposition rate, and adhesion to the substrate on the morphology of polycrystalline thin films. The simulations provide insight into the dominant microscopic mechanisms that drive the structure evolution during thin film processing.

Parallel Continuous-Time

A parallel multibit continuous-time (CT) DeltaSigma analog-to-digital converter for an orthogonal-frequency-division-multiplexing (OFDM) ultrawideband receiver intended to operate according to the IEEE 802.15.3a or the ECMA 368 (ISO/IEC 26907) standards has been designed. The overall CT DeltaSigma converter consists of two modulators covering two unequal subbands (low-pass (LP) and bandpass (BP) subbands) that are arranged to operate in parallel and whose respective noise transfer functions (NTFs) are designed to match its corresponding frequency band. The composite NTF for the overall converter is defined as the minimum gain value out of these two individual NTFs. The LP and BP subbands were designed by using third- and fourth-order modulators, respectively, based on a 3-bit quantizer and operating at a clock frequency of 1056 MHz. NTF zero locations were optimized according to the criterion that all the in-band composite NTF gain maxima have approximately the same value. Combining OFDM signal characteristics and converter parameters, the effect of the quantization noise on the overall converter performance has been analytically derived. A simulation program has been realized to verify the performance of the converter.

\Delta\Sigma

In this work the design of a continuous-time @D@S modulator for Gigabit Ethernet applications is presented. The input bandwidth and oversampling ratio are, respectively, 62.5MHz and 8, resulting in a clock frequency of 1GHz. It was designed and implemented in a standard 90nm CMOS technology. The active area of the modulator measures 0.0207mm^2. It consists of a loop filter based on RC-opamp integrators and a 3-bit quantizer which includes a data weighted averaging scrambler. A digital tuning scheme to deal with process variations has also been included. System level simulations including several non-ideal effects have been carried out in order to determine in detail the performance of the converter. Experimental results show a resolution of 7.1 effective bits, and a power consumption of 10.8mW from a nominal power supply of 1V.

ADC for OFDM UWB Receivers

A detailed analysis of the impact of a hysteretic quantizer on a multibit, Sigma-Delta modulator has been carried out in this paper. Both discrete-time and continuous-time modulators have been considered. A qualitative modeling of the hysteretic quantizer based on a hysteretic block followed by an ideal quantizer was proposed. Due to the hysteresis effect, the quantizer output signal is delayed and distorted with respect to the quantizer input signal, where the delay causes a phase-shift independent on the signal frequency. Yet, the effect of the hysteresis depends on the input signal amplitude. This model was validated by using system-level simulations for a second order, 3-bit, discrete-time Sigma-Delta modulator. A linear model for hysteresis was derived by assuming a narrow hysteresis cycle. The quantizer input signal plays a fundamental role in the discussion. In order to include this signal into the linear analysis some approximations are proposed. The quantizer output signal is decomposed by the use of the Fourier series analysis only into the in-phase and quadrature components (with respect to the input signal) whose Fourier series coefficients can be analytically calculated. A quantitative analysis for both a second order, 3-bit, DT and CT Sigma-Delta modulators including a hysteretic quantizer was carried out. For the CT modulator, finite GBW in amplifiers, excess loop delay, and a hysteretic quantizer were considered separately and combined. A good agreement with both system-level simulations and experimental results is found, despite the approximations considered for the quantizer input signal.

A 1-GHz, multibit, continuous-time, delta-sigma ADC for Gigabit Ethernet

Abstract As integrated circuit devices scale into the deep sub-micron regime, ion implantation will continue to be the primary means of introducing dopant atoms into silicon. Different types of impurity profiles such as ultra-shallow profiles and retrograde profiles are necessary for deep submicron devices in order to realize the desired device performance. A new algorithm to reduce the statistical noise in three-dimensional ion implant simulations both in the lateral and shallow/deep regions of the profile is presented. The computational effort in BCA Monte Carlo ion implant simulation is also reduced.

Detailed analysis of the effect of a hysteretic quantizer in a multibit, Sigma-Delta modulator

The amorphization of silicon due to atomic displacement during ion implantation has been simulated. A model based on Monte Carlo calculation reproduces very well the depth profile of atomic mixing and displacement length of host silicon atoms reported by previous experiments. The critical displacement in the depth direction for amorphization has been determined to be 5 A. This average threshold value is shown to be universal for identification of amorphous regions in silicon for a wide range of implantation conditions involving different doping species, acceleration energies, and doses.

Algorithm for statistical noise reduction in three-dimensional ion implant simulations

A statistical 3D damage accumulation model, based on the modified Kinchin–Pease formula, for ion implant simulation has been included in our physically based ion implantation code. It has only one fitting parameter for electronic stopping and uses 3D electron density distributions for different types of targets including compound semiconductors. Also, a statistical noise reduction mechanism based on the dose division is used. The model has been adapted to be run under parallel execution in order to speed up the calculation in 3D structures. Sequential ion implantation has been modelled including previous damage profiles. It can also simulate the implantation of molecular and cluster projectiles. Comparisons of simulated doping profiles with experimental SIMS profiles are presented. Also comparisons between simulated amorphization and experimental RBS profiles are shown. An analysis of sequential versus parallel processing is provided.

Monte Carlo simulation of silicon atomic displacement and amorphization induced by ion implantation

Abstract An efficient binary collision approximation ion implant code with enhanced prediction capabilities is presented. It includes recent improvements in physical models for compound semiconductors. It uses only one fitting parameter for low dose implantations. A periodic ab initio full bond electron density for the target is used. Damage accumulation is supported using a modified Kinchin–Pease model. To speed-up the code a refined algorithm for statistical noise reduction is also included in a three-dimensional case, including the lateral and shallow zones. The agreement with experiments is good for different target materials. A comparison with experimental SIMS results for several projectiles and targets is presented.

Statistical 3D damage accumulation model for ion implant simulators

A comprehensive analysis of the impact of the finite gain-bandwidth product (GBW) in amplifiers and the excess loop delay on a parallel multibit CT ΔΣ ADC for an orthogonal frequency division multi...